A fork in the road for public communication?
2026 may mark a tipping point for digital public communication. On the one hand, generative AI is rapidly embedding itself in our processes, promising unprecedented efficiency. On the other, a growing demand is emerging to regain control over our tools, data and sovereignty. How do we strike the right balance between these two forces? And what if this tension were not a problem, but rather a solution for the years ahead?
By Marc Cervennansky, head of the web and social media centre at Bordeaux Métropole.
The year has barely begun and yet specialists in digital public communication (1) are already exchanging views on current and emerging trends as they prepare the programme for the forthcoming National Meetings on Digital Communication to be held in Rennes in September.
Rest assured, I won’t be revealing the programme, which in any case is still being finalised. However, interesting discussions with my peers, which in my view reveal two major trends, have inspired this article.
The surge: AI as the new standard
I’ve noticed that in my daily conversations, in less than a year, the question has shifted from “Should we use AI?” to “How can we integrate it ethically and responsibly?” The local authorities who consult me no longer ask whether they should use AI, but which tools to choose, how to train their teams and what level of transparency to display. Some are drafting ethical charters, organising workshops tailored to different job categories, or negotiating specific clauses in public procurement contracts.
At the same time, GEO (Generative Engine Optimization) is establishing itself as the new SEO. Writing for search engine optimisation alone is no longer sufficient: content now needs to be “AI-compatible,” anticipating how it will be processed, reformulated and perhaps distorted by the AIs that respond in place of our websites.
This race can provoke anxiety and a the fear of standardisation. If everyone uses the same tools to produce content optimised for the same algorithms, what remains of our uniqueness? Of our ability to embody a region or maintain a distinctive public voice? And beyond that, what visibility remains for public services, swallowed and digested in AI?
The response: sovereignty, restraint and authenticity
Faced with this tsunami, largely driven by GAFAM, a counter-offensive is beginning to emerge. Not a rejection of technology, but a desire to take back control.
First focus: digital sovereignty. The tense geopolitical context makes dependence on GAFAM – especially Microsoft – increasingly problematic for the public sector. Yet national initiatives are emerging, alongside local solutions. The LLM projects in Brittany and Corsica are far from mere cultural gimmicks: they test our ability to create tools that reflect our values, are hosted locally and self-governed.
Second focus: digital sobriety. I’ve noticed a curious paradox: even though we’ve been talking for years about carbon footprints and responsible digital practices, the advent of generative AI, which is extremely energy-hungry, has pushed these concerns aside. But some local authorities are already experimenting with ‘de-digitisation’, the right to be non-digital and other concrete solutions for a more sustainable digital future. How do we make sure we stay on this path?
Third focus: authenticity. In response to the ‘editorial sludge’ created by AI, a counter-trend is emerging in the form of homemade web content, deliberate imperfection and artisanal formats. On social media, the most engaging content isn’t the slickest, but rather the most genuine and personal. Some cities are opting to feature identifiable faces in videos to deliver their messages, rather than relying entirely on AI-generated content.
The question is not “for or against?’ but “under what terms?”
This is where we stand: caught between the urgency to use AI to stay in the race and the need not to hand over control to tools we don’t fully understand or manage. Between the promise of efficiency and the risk of losing our sense of purpose.
I believe this tension is actually healthy. It forces us to move beyond both techno-enthusiastic naivety and paralysing fear. It pushes us to ask the right questions: AI should act as an assistant, not a replacement. Tools must be selected according to our ethical values, not simply based on cost. Editorial governance across all media should make clear decisions, instead of being carried by the trend.
A fork in the road lies ahead. It’s a choice about taking control. Control over our tools, our data and our messages. Regaining control over what lies at the heart of our work, namely building connections, informing with integrity and serving the public interest.
The coming months will be crucial. Let’s talk about it again in Rennes in September.
(1) This article draws on ideas from Franck Confino, Pierre Bergmiller, Estelle Soleillant, Inès Slama, Philippe Couve, Lauric Didier-Mougin, Marie-Claire Taché, Patrice Razet, Julien Bordas, Anne-Claire Dubreuil, Maxime Goguet and myself.
